0 II
H I C-R
Ia. R = C H 3
IIa. R = C H J
Ib. R = -CH2
IIb. R = - C H *
H
mCyR I
c=o
Huisgen and Rauenbusch' reported that the catalytic reduction of I a does not proceed beyond the hydroxy stage (IIa) which then cyclizes into the lactone (IIIa), which is not easily amenable to further reduction. \JTe found that the desired reduction can be easily accomplished by working in neutral or slightly alkaline medium, this means in fact the reduction of a salt of I a instead of I a itself. The formation of the lactone (IIIa) is thus prerented. JT7hen I a in the form of its sodium salt in water is hydrogenated with palladium on charcoal catalyst at 80' and a pressure of 40 p.s.i. two moles of hydrogen are absorbed in eight hours to afford IVa. The results are similar with 2-phenylacetylbenzoic acid (Ib). The hydroxy acid (IIb) can be also reduced into IVb under analogous conditions. The yields are in the range of 83-92%. Experimental 2-Ethylbenzoic Acid (IVa).--4 16.4-g. sample of oacetylbenzoic acid (Ia; 0.100 mole; from Aldrich Chemical Co.) was dissolved in a mixture of 100 ml. of 1.0 N sodium hydroxide solution and 70 ml. of water (pH 7). It wa8 hydrogenated after the addition of 10 g. of Pd-Darco catalyst ( l O 7 , palladium) at 40 p.s.i. starting pressure. After the uptake of 0.1 mole of hydrogen in 2 hr. further uptake ceased, therefore it was heated to 80' and hydrogenated for 6 hr. a t 40 p.s.i. until another 0.1 mole of hydrogen was consumed. After filtration through Supercel and acidifying with hydrochloric acid 12.4 g. of IVa was obtained (83% yield) m.p. 62-63' (lit.,l m.p. 60-62"). ,4nal. Calcd. for CsHlo02: C, 71.9; H, 6.75. Found: C, 72.02; H, 6.05. Equiv. wt. (by titration) 149; theory 150.08. 2-Phenylethylbenzoic Acid (1%). A. From 2-Phenylacetylbenzoic Acid (Ib).-A 22.2-g. sample of 3-benzalphthalide (0.100 mole) was transformed into the sodium salt of 2-phenylacetylbenzoic acid by saponification with a solution of 5 g. of sodium hydroxide in 100 ml. of water (20 min. a t 90-95'). Fifty milliliters of water was added and the pFI of the solution adjusted to 8.5 by the addition of hydrochloric acid. After cooling to room temperature the solution was filtered to remove a small amount of by-product; it was then hydroqenated after the addition of 7 g. of Pd-Darco catalyst (10% palladium) at 80' with a starting pressure of 40 p.s.i. In 8 hr. 0.2 mole of hydrogen was taken up and the uptake practically ceased. Acidification of the filtered solution gave 20.71 g. of IVh, m.p. 128-130" (lit., 130-131°; niixed m.p., the same; over-all yield 91.6%.
B. From 2-( 1-Hydroxy-2-phenylethy1)benzoic Acid (IIb). -A solution of the sodium salt of 2-( 1-hydroxy-2-phenylethy1)benzoic acid was prepared by saponifying 22.4 g. (0.100 mole) of 3-benzylphthlide3 by refluxing for a short time with 5 g. of sodium hydroxide in 150 ml. of water. The solution of I I b then obtained was buffered to pH 8-9 by the addition of 14.2 g. of disodiiim phosphate and 1.1 ml. of 85% phosphoric acid. It was hydrogenated for 8 h r . a t 120' a t 40 p.s.i. with 3.5 g. I'd-Darco catalyst (5% pallndiun). By acidification of the filtrate 19.7 g. of 2-phenylethylbenzoic acid was obtained (85% of theory); m.p. 125-128"; did not depress the melting point of an authentic sample of IVb. (3) S. Natilson and S. P. Gottfried, J. Am. Chem. Soc., 68, 1432 (1936).
A Reinvestigation of the Hydrogenolysis of Hydroxyl Groups of the Stereoisomeric 3-Phenylcholes tanols EDGAR
11'.
GARBISCH,
JR.'
School of Chemistry, L7nivwsity of Jfinnssota, Minneapclis, Xinnesotn
Received April 16. 1962
Recently Zderic, Rivera, and Lim6n found that Raney nickel hydrogenolysis of the hydroxyl groups of the epimeric 3-phenylcholestanols led to the common product, 3p-phenylcholestane.2 Reportedly, the Raney nickel hydrogenolysis of 3aphenylcholestan-3~-01 was the first inst'ance in which a hydroxyl group was reduced with inversion of c~nfiguration.~A mechanism was proposed.2 Observations from work now underway have led to a reinvestigation of the hydrogenolysis of the 3-phenylcholestanol system.2 The series of experiments summarized below demonstrate that W-2 Raney nickel hydrogenolysis of the hydroxyl groups of the st'ereoisomeric 3-phenylcholestanols indeed proceeded wit,h a high degree of retent'ion of configuration (probably > goyo), and that 3aphenylcholestane was rapidly equilibrated to the t hemodynamically more stable Sp-phenylcholest,aneby Raney nickel in refluxing ethanol. In accordance with earlier reports,2 the treatment of 3cr-phenylcholestan-3~-01(I) with Raney nickel in refluxing ethanol led to 3p-phenylcholestane (IV). Likewise, the hydroxyl group of 3pphenylcholestan-3a-01 (11) was hydrogenolyzed to give pure IV.4 The proton magnetic resonance spectrum of IV exhibits absorptions a t 7.587 and 2.897 for the C-3 proton and phenyl protons, re(1) N. s. F . Postdoctoral Fellow, 1961-1962. Present address: Department of Chemistry, University of Chicago, Chicago 37, Illinois. (2) J. A. Zderic, Ma. E. C. Rivera, and D. C. Limb, J. Am. Chem. Sor., 82, 6373 (1960). (3) See Ref. 1 for background references. See a19o: D. Cram and J. Allinaer, ibid., 76, 4516 (1854): D. Curtin and S.Schrnukler. ibid.. 77, 1105 (1955): and S. Mitsui and S. Imaizumi. Bull. Chem. Sor. Japan, SI, 774 (1961). (4) As little as 5 % of I11 in IV or vice versa could be detected reliably in the infrared using the absorptions between 13.1 p and 14.3 p.
3364
NOTES
~pectively.~However, when I was stirred with Raney nickel a t room temperature, pure 3a-
&; -+$
HO
Ph
Ph
I
I11
I
OH
H I1
IV
phenylcholestane (111) was ~ b t a i n e d . The ~ proton magnetic resonance spectrum of I11 exhibits absorptions a t 6.95, and 2.807 for the C-3 proton and phenyl protons, respectively. Under refluxing ethanol hydrog6nolysis conditions, I11 was gradually converted to IT'. Hydrogenolysis of I1 proceeded slowly at room temperature (as compared with I) to give IV. Under comparable conditions, I was converted to I11 with < 5% of IV. By modifying the catalyst through the addition of small amounts of sodium methoxide, the rate of equilibration was apparently reduced relative to the rate of hydrogenolysis, and higher temperatures could be employed to further demonstrate the stereospecificity of the hydrogenolysis which nom proceeded a t a convenient rate. Using this procedure, I mas hydrogeiiolyzed to a mixture of 85% of I11 and 15% of IT' in refluxing ethanolR m e y nickel. Likewise, I1 was converted to pure 1v.4
Experimental 3a-Phenylcholestan-3p-ol ( I ) and 3fi-phenylcholestan-3a01 (11) were prepared as described previously2 and melted a t 109.5-110" and 164-165", respectively; reported,2 m.p. 100-103" and 163-165", respectively. Hydrogenolysis at 7S0.--A mixture of 50 mg. of I , 2 g. of W-26 Raney nickel and 4 ml. of ethanol was refluxed for 20 min. The ethanol solution then was decanted into a mixture of 10 ml. of pentane and 10 ml. of water. The catalyst was washed twice with 10-ml. portions of pentane and the washings were added to the water-pentane mixture. The pentane laver was washed with water and the pentane then evaporated. Infrared analysis of the residue (37 mg.) gave approximately 85% of IV and 15% of III.4 After one recrystallization from acetone, pure 36-phenylcholestane (IT), m.p. 113-114.5', was obtained; infrared spectrum ( C S I ) ,13.16 (s) p and 14.25 (s) p ; reported,2 m.p. 113-114". The hydrogenolrsis of I1 as described above led to 42 nig. of pure ITT.4 Hydrogenolysis at Room Temperature.--A mixture of 100 mg. of I, 3 g. of Raney nickel, and 10 ml. of ethanol
_____
( 5 ) The proton niaunetlc resonance npwtra urir determined by Dr
W. R. Schwabarher. (6) R. Mozingo, O ~ Q . Syntheses, %+ol. 111, John Wiley & Sons, Inc., Neu Pork, 181 (1956). The Raney niakel uaed En this m o r k wae freahly prepared and of the same batch.
T'OI,.
27
was stirred for 10 min. a t room temperature. The reaction mixture was processed as previously described to give 84 mg. of pure 3a-phenylcholestane (111), m.p. 82-85°.4 Several recrystallizations from acetone raised the melting point to 86.5-88.5'; infrared spectrum (CS?), 18.51 (m) p , 13.7'7 (m) p , and 14.25 (s) p . Anal. Calcd. for C33H62: C, 88.32; H, 11.68. Found: C, 88.51, H, 11.48. A mixture of 40 mg. of 111, 4 ml. of ethanol, and 3 g. of
Raney nickel was refluxed for 15 min. The crude product of this reaction analyzed for 35% of I11 and 657' of 11(infrared). .i mixture of 30 mg. of 11, 1 nil. of ethanol, and 3 g. of Haney nickel was stirred for 2 hr. The infrared spectrum of the rrude hydrogenolysis product showed about a 50c/, conversion of I1 to IV and no detectable amount of 111.' A comparable treatment of I led to I11 contaminated with < 5% of IV. Hydrogenolysis at 78' with Added Sodium Methoxide.Treatment of 50 mg. of I with 2 g. of Raney Nickel and 10 mg. of sodium methoxide in 4 ml. of ethanol a t reflux for 7' min. led to the isolation of 36 mg. of a mixture of 85y0 of I11 and 15% of IV (by infrared analysis). The treatment of 50 mg. of I1 in the same manner a8 above led to 37 mg. of pure ITT.
Synthesis of 3-Substituted 3-Phenylphthalides FRANKN. JONESAND CHARLES R. HAUSER
Department of Chemistry, Duke University, Durham, North Carolina Received April 16, 1963
In connection with other work, we have found that the pseudo chloride of o-benzoylbenzoic acid, 3-chloro-3-phenylphthalide (I), reacts with organocadmium reagents to give 3-substituted 3-phenylphthalides in excellent yields (equation 1). The reaction is effected by the technique recommended by Cason2 for preparation of ketones from acid chlorides.
I
0
(1)
11.R = CHs III.R=n-C4H9 IV R=CsHs
Three examples of the reaction were effected. Treatment of I with dimethylcadmium, prepared from methylmagnesium bromide, gave 3-metliyl-3phenylphthalide (11) in 88% yield, based on obenzoylbenzoic acid. Similarly, di-n-butylcadmium (1) Supported b y the Army Research Office (Dnrham). (2) J. Cason, Chsm. Rev., 40, 15 (1947).
